72 BENJ. PIKE'S, JR., DESCRIPTIVE CATALOGUE. 



is at on the arc, and the optical axis of the telescope must 

 be parallel to the plane of the instrument. We shall speak 

 separately of each of these adjustments. 



To examine the adjustment of the Index-glass. Move the 

 index forward to about the middle of the limb, then, hold- 

 ing the instrument horizontally with the divided limb from 

 the observer, and the index-glass to the eye, look obliquely 

 down the glass, so as to see the circular arc, by direct view 

 and by reflection, in the glass at the same time ; and if they 

 appear as one continued arc of a circle, the index-glass is 

 in adjustment. If it requires correcting, the arc will appear 

 broken where the reflected and direct parts of the limb 

 meet. This, in a well-made instrument, is seldom the case, 

 unless the sextant has been exposed to rough treatment. 

 As the glass is in the first instance set right by the maker, 

 and firmly fixed in its place, its position is not liable to alter, 

 therefore no* direct means are supplied for its adjustment. 



To examine the Horizon-glass, and set it perpendicular to 

 the Plane of the Sextant. The position of this glass is 

 known to be right, when by a sweep with the index, the 

 reflected image of any object passes exactly over or covers 

 its image, as seen directly ; and any error is easily rectified 

 by turning the small screw, i, at the lower end of the frame 

 of the glass. 



To examine the Parallelism of the Planes of the two 

 Glasses, when the Index is set to Zero. This is easily ascer- 

 tained ; for, after setting the zero on the index to zero on 

 the limb, if you direct your view to some object, the sun 

 for instance, you will see that the two images (one seen by 

 direct vision through the unsilvered part of the horizon- 

 glass, and the other reflected from the silvered part) coin- 

 cide or appear as one, if the glasses are correctly parallel to 

 each other ; but if the two images do not coincide, the 

 quantity of their deviation constitutes what is called the 

 index error. The effect of this error on an angle measured 

 by the instrument is exactly equal to the error itself; there- 

 fore, in modern instruments, there are seldom any means 

 applied for its correction, it being considered preferable to 

 determine its amount previous to observing, or immediately 

 after, and apply it with its proper sign to each observation. 

 The amount of the index error may be found in the follow- 

 ing manner : clamp the index at about 30 minutes to the 

 left of zero, and looking towards the sun, the two images 



